The present invention relates to a method of discriminating the attribute of an image, an image processing apparatus adopting the discrimination method and an image forming apparatus.
Recently, when an image incorporating characters and gradation images mixed therein is treated, recording for obtaining a hard copy is an issue that needs to be resolved. An electrophotographic method is in common use as a method of recording a digital image. The foregoing method has unsatisfactory performance of expressing a density of two to several levels per picture point to be recorded. Therefore, when a gradation image is expressed, a pulse width modulation method or the like must be employed. The pulse width modulation method is broadly divided into a one-pixel modulation method and a two-pixel modulation method in terms of the period of pulses in the pulse width modulation. Although the former method is able to clearly record characters because of excellent resolution, the foregoing method is of inferior gradation expressing characteristic. On the other hand, the latter method excels in gradation expressing characteristic which enables a photograph or the like to smoothly be reproduced. However, the latter method is of inferior resolution. Although the modulation methods are able to realize either of the resolution and the gradation expressing characteristic, both of the foregoing requirements cannot simultaneously be realized when a recording operation is performed.
To record an image such that both of the resolution and the gradation expressing characteristic are realized, an image field discrimination process is performed. That is, an image which must be recorded is discriminated into fields, for example, photographs, in which the gradation expressing characteristic is of importance and fields, for example, characters or line drawings, in which the resolution is of importance. In accordance with a result of the discrimination, the recording method is switched.
As an image field discrimination method, a method is known in which the difference in change in local densities between gradation fields and character fields or the difference in the local patterns is used. As an example of the former method, a method has been disclosed in Jpn. Pat. Appln. KOKAI Publication No. 58-3374. The method has steps of dividing an image into small blocks, and calculating the difference between a highest density and a lowest density in each block. Moreover, if the difference is larger than a threshold value, a discrimination is made that the subject block is a character image field. If the difference is smaller than the threshold value, a discrimination is made that the subject block is a gradation image field. The above-mentioned method is able to perform accurate discrimination if the image is composed of only continuous gradation images, such as photographs, and characters. However, there arises a problem of unsatisfactory discrimination accuracy in a field, for example, a dot image, in which local density change frequently occurs. There arises another problem in that a gradation image having a sharp edge thereon is incorrectly discriminated as a character field. As an example of the latter method, a method has been disclosed in Jpn. Pat. Appln. KOKAI Publication No. 60-204177. The foregoing method has steps of Laplacian-filtering an image; binarizing the image; and performing discrimination in accordance with a shape of, for example, a 4xc3x974 pattern. The above-mentioned method is able to discriminate even a dot image. However, also the above-mentioned method had a problem in that an edge portion on a gradation image can frequently incorrectly be discriminated as a character image.
If the foregoing methods are combined with each other or if a method is furthermore employed in which correction is performed in accordance with results of discrimination of surrounding pixels by using a characteristic that an image field is constant over a somewhat wide area, the discrimination accuracy can be improved. However, the scale of the circuit results in limitation of reference fields to several pixels. Thus, a satisfactory discrimination accuracy cannot be realized.
As described above, the method using information of the local density among the conventional image field discrimination methods suffers from a problem in that discrimination accuracy of an edge portion of a gradation image having a micro structure similar to that of a character and a rough dot image portion deteriorates. Another problem arises in that a character formed by thick lines and a portion in lines cannot easily be discriminated as a character.
The known method of macroscopically analyzing the structure of a document image, which is able to accurately discriminate existence of a character or a gradation image, suffers from unsatisfactory position-resolving power. Thus, accurate discrimination in pixel unit cannot easily be performed.
An object of the present invention is to provide an image field discrimination method which is capable of realizing both excellent discrimination accuracy and excellent position resolving power and with which a portion in a characteristic can correctly be discriminated, an image processing method using the same and an image forming apparatus.
According to one aspect of the present invention, there is provided an image processing apparatus comprising: the field separating means for separating an original image into plural types of fields in response to a first image signal obtained at a first density of the supplied original image; characteristic value calculating means for calculating a characteristic value of the original image in response to a second image signal of the original image obtained at a second density which is higher than the first density; discrimination means for discriminating an image field of the original image in accordance with the characteristic value calculated by the characteristic value calculating means to correspond to the type of the field separated by the field separating means; and an image processing means for processing a predetermined image processes corresponding to a result of discrimination of the image field performed by the discrimination means, on the second image signal.
The present invention having the above-mentioned structure is different from a conventional method in which discrimination of an image field is performed by one time. The present invention has the structure that discrimination of a field in response to a rough signal of an original image is performed. Moreover, a characteristic value indicated by a dense signal of the original image is obtained. Then, final discrimination of the field is, in each field indicated by the rough signal, performed in accordance with the characteristic value. As a result, further accurate discrimination of a field can be performed.
That is, the conventional method has the structure that discrimination of a field is performed only one time in response to a dense signal. Thus, incorrect discrimination of an edge in a photograph field as a character field can easily be made. However, the present invention arranged to first perform macro field discrimination in accordance with the rough signal is able to prevent incorrect discrimination that an edge in the foregoing photograph field is a character. Since the second process is performed such that the field is discriminated in accordance with the characteristic value indicated by the dense signal, a micro character field in a character field and a background field are discriminated from each other. Thus, the character field and the background field can individually be detected. Since high contrast similar to that of a character field is not provided for the background field, that is, since low contrast is provided for the background field, generation of noise on the background can be prevented.
Therefore, the field discrimination according to the present invention and arranged to perform two steps is able to prevent incorrect discrimination between a photograph field and a character field, detect a background field in a character field and thus realize a background field free from any noise.
Also a method according to the present invention enables both of a character field and a gradation field to satisfactorily accurately be discriminated from each other because of the above-mentioned reason.
The discrimination of an image field according to the present invention can favorably be employed in a digital color image forming apparatus. In this case, accurate discrimination of an image field can be performed and an image in a satisfactory state can be formed.
Additional objects and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objects and advantages of the invention may be realized and obtained by means of the instrumentalities and combinations particularly pointed out hereinafter.